Skip to main content
Visual component Add web search and page fetch to your Langflow flows with two components, Keenable Search and Keenable Fetch. They ship in the lfx-keenable extension bundle, run without an API key, and work as tools for a Langflow agent. → lfx-keenable on PyPI · source

Install

Install the package into the same Python environment as Langflow. It needs Python 3.10 or later and works with lfx 1.10 and later 1.x releases:

Restart Langflow

Restart the Langflow server. To use your API key for every component, set it in the environment first:
The Keenable components appear in the palette under the keenable group.

Use the components

Drag Keenable Search or Keenable Fetch onto the canvas. To give them to an agent, turn on Tool Mode in the component’s header menu, then connect its Toolset port to the Agent component’s Tools port. In Tool Mode, the model fills Search Query, URL, Site and the date inputs per call. Keenable Search outputs a Results table with one row per result: title, url, description, snippet, published_at and acquired_at. Keenable Fetch outputs a Page table with one row: the page content as markdown, plus title, description, author and published_at when the page has them. KEENABLE_API_URL overrides the API address, for example to route calls through a proxy. It must be an https:// URL. No component input sets it.

Verify

Add Keenable Search to a new flow, type a query into Search Query and run the component. The Results output lists the pages it found. If the table has a single error column instead, read the message and check the table below.

Troubleshooting

The components catch errors and return one row with an error field, so the rest of the flow keeps running. For the same tools in other visual builders, see Dify and n8n.
Set KEENABLE_API_KEY as an environment variable, or set the key on the component input, to remove the hourly request cap. If you don’t set a key, calls run on the shared keyless tier at lower rate limits.

RAGFlow

Give RAGFlow agents web search with the Keenable tool built into RAGFlow’s visual agent builder. Works without an API key.

Convex

Add Keenable search and fetch to a Convex app with the @keenable/convex component, with or without an API key.

Replit

Give Replit Agent web search and page fetch through the Keenable MCP server: one click, then sign in to Keenable.

Connect your agent (MCP)

Connect the Keenable MCP server to Claude Code, Cursor, Codex and other clients for web search and fetch with no key. Sign in for higher limits and Monitors.